NTIRE 2025 Challenge on Cross-Domain Few-Shot Object Detection: Methods and Results (2025-04-14T00:00:00.000000Z)

TL;DR

An overview of the 1st NTIRE 2025 CD-FSOD Challenge is presented, highlighting the proposed solutions and summarizing the results submitted by the participants.

Abstract

Cross-Domain Few-Shot Object Detection (CD-FSOD) poses significant challenges to existing object detection and few-shot detection models when applied across domains. In conjunction with NTIRE 2025, we organized the 1st CDFSOD Challenge, aiming to advance the performance of current object detectors on entirely novel target domains with only limited labeled data. The challenge attracted 152 registered participants, received submissions from 42 teams, and concluded with 13 teams making valid final submissions. Participants approached the task from diverse perspectives, proposing novel models that achieved new state-of-the-art (SOTA) results under both open-source and closed-source settings. In this report, we present an overview of the 1st NTIRE 2025 CD-FSOD Challenge, highlighting the proposed solutions and summarizing the results submitted by the participants.

Authors

References44 items

NTIRE 2025 Ambient Lighting Normalization Challenge Report

NTIRE 2025 Challenge on Raw Image Restoration and Super-Resolution

TIRE 2025 challenge on Text to Image Generation Model Quality Assessment

NTIRE 2025 Challenge on UGC Video Enhancement: Methods and Results

NTIRE 2025 Challenge on Image Super-Resolution ($\times 4$): Methods and Results

Qwen2.5-VL Technical Report

Locate Anything on Earth: Advancing Open-Vocabulary Object Detection for Remote Sensing Community

Unified View Empirical Study for Large Pretrained Model on Cross-Domain Few-Shot Learning

The Solution for CVPR2024 Foundational Few-Shot Object Detection Challenge

Sharing Key Semantics in Transformer Makes Efficient Image Restoration

Grounding DINO 1.5: Advance the "Edge" of Open-Set Object Detection

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

The Ninth NTIRE 2024 Efficient Super-Resolution Challenge Report

OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models

Mixtral of Experts

Improved Region Proposal Network for Enhanced Few-Shot Object Detection

DINOv2: Learning Robust Visual Features without Supervision

Free-Lunch for Cross-Domain Few-Shot Learning: Style-Aware Episodic Training with Robust Contrastive Learning

TGDM: Target Guided Dynamic Mixup for Cross-Domain Few-Shot Learning

ME-D2N: Multi-Expert Domain Decompositional Network for Cross-Domain Few-Shot Learning

Learning attention-guided pyramidal features for few-shot fine-grained recognition

ELEVATER: A Benchmark and Toolkit for Evaluating Language-Augmented Visual Models

Underwater Species Detection using Channel Sharpening Attention

LoRA: Low-Rank Adaptation of Large Language Models

MDETR - Modulated Detection for End-to-End Multi-Modal Understanding

Learning Transferable Visual Models From Natural Language Supervision

Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity

A realistic fish-habitat dataset to evaluate algorithms for underwater visual analysis

Adversarial Weight Perturbation Helps Robust Generalization

Frustratingly Simple Few-Shot Object Detection

Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation

HuggingFace's Transformers: State-of-the-art Natural Language Processing

Objects365: A Large-Scale, High-Quality Dataset for Object Detection

Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark

The Open Images Dataset V4

Drone-Based Object Counting by Spatially Regularized Regional Proposal Network

DeepFruits: A Fruit Detection System Using Deep Neural Networks

Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations

A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects

Im2Text: Describing Images Using 1 Million Captioned Photographs

A Survey on Mixture of Experts

Detect everything with few examples

Meta-FDMixup

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding